Thunderstone Support Forums

I ran a walk with verbosity set at 4 and one of the messages was the following any ideas what this means?

Had this error: Reply size 175 != Content-Length 160 (truncated) for http://www.jan.wvu.edu/sbses

Counting challenged web server software. It said it was about to send 160 bytes then proceeded to send 175.

This error means the web server said the content will be 160 bytes (Content-Length header), but sent 175 bytes of actual data instead, so the client (Webinator) can't be sure if the page is truncated or not. It indicates an error in the web server or CGI program serving that URL.

In this case, the issue seems to be the redirect that the server is issuing for /sbses to /sbses/. The last line of the response is `</BODY></HTML>' which is 15 characters including the newline; my guess is that something adds that line *after* the Content-Length is computed.

Check your web server config for how directory redirects are handled. Also, change references to /sbses to /sbses/ (the correct URL) to avoid the redirect in the first place.

I had fixed this problem by modifying the search script. As I recall we commented out something in the search script. I can't remember what the fix was. Please help. Several hundred docement are nolonger in my index. thanks

It looks like a new walk that just finished indexed the documents. However I am trying to find out what would cause this. A walk just this morning just had "unexpected errors" and so the problem documents didn't get indexed. I just ran a refresh walk and now these 350 documents (all from the same site) are now in the index.

In dowalk, not search, add
"Reply size"
to the list of $partialokpageexprs .

I am having additional problems because I am now getting this error. If I do a refresh rewalk manually the "reply size" error is there and the docs get indexed. If it is a scheduled refresh walk I get the "Unexpected content..." error and the docs aren't indexed.

Can I add this error to dowalk also or will this action cause additional unintended consequences?

This is really weird that auto walks fail and manual walks work.

The link : http://www.access-board.gov/play/guide/summary.htm
Had this error: Unexpected content (61 bytes) for http://www.access-board.gov/play/guide/summary.htm
Referenced by : http://www.access-board.gov/play/guide/intro.htm
http://www.access-board.gov/play/guide/terms.htm

Yes, add "Unexpected content" to the list.

Make sure you only have one version of texis and dowalk installed. All I could think of is that the schedule is running a different version of something than is running from the CGI environment of the webserver (manual walk).